The development of communication networks and the need for high data rate transmission led to the creation of fiber optic networks. The increasing expansion of computer networks highlighted the need to transmit a large volume of data. One of the challenges affecting elastic fiber optic networks is the optimal use of available resources. The reason is that the increased number of users disrupts network services. One way to optimize elastic fiber optic networks is to find appropriate routing for message broadcasting. Reliability and security criteria have not been discussed in the research literature. This study aims to use an efficient algorithm for routing and spectrum or frequency allocation. To enhance security and reliability, low-latency and reliable alternative paths are introduced for each node in the network so that the relevant objectives can be pursued even after the node fails. The proposed solution was evaluated on a complex problem and compared with the results of particle swarm optimization and genetic algorithms. When comparing the algorithms, criteria such as best-case convergence, average convergence, stability, and some statistical parameters were considered. The implementation conditions of all three algorithms were considered completely equal and fair. The results show the correct and appropriate performance of the global competition algorithm compared to other methods.